One Down, One to Go

There was much rejoicing among analysts, economic forecasters and financial market participants on June 6 when the BLS told us that total nonfarm payroll employment on a seasonally adjusted basis set a new record in May 2014 with 138,463,000 such jobs then. The old record of 138,350,000 such jobs on a seasonally adjusted basis was set in January 2008, which was the first full month of the 18-month long recession that began in December 2007 and ended in June 2009.

The chart shows the pattern of this widely followed economic series since January 1978. It is quite obvious that instead of the fairly quick recovery in such jobs that followed every recession from 1945 to the 1981-1982 one, the length of time to return to previous levels has gotten longer and longer with every recession beginning with the 1990-1991 event.


While many reports on this new record contained statements claiming that all the jobs lost in the recession had been made up, that is not technically true. What IS true is that the total number of jobs has now been matched. But tens of millions of actual jobs that disappeared in 2008-2010 will never come back. They have just been replaced by other jobs.

In addition to that, the total population and the labor force have grown a lot over this time frame. Some estimates are that we might need as many as five million more jobs today just to be even with how well off we were in January 2008 in terms of payroll employment.

It turns out that the pattern of nonfarm payroll jobs today is vastly different from what it was back in January 2008. Here are some of the comparisons.

By far the largest number of net new nonfarm payroll jobs over that period is found in the “health care and social assistance” category, which has risen by 2,150,000 such jobs. Next is “Accommodations and food services” with an increase of 941,000. “Professional and technical services” jobs have grown by 512,000. “Education services “has gained 425,000 jobs and “Temporary help services” has added 307,000 jobs since January 2008.

Not very surprisingly, the biggest loser is jobs in manufacturing. There were 1,650,000 fewer of those in May 2014 than in January 2008. This is hardly a new story. The peak was 19,553,000 jobs back in June 1979. The recent trough counted 11,453,000 such jobs, which was the lowest number since March 1941, well before the US became involved in World War II. The May 2014 level of 12,099,000 is still lower than in June 1941, both on a seasonally adjusted basis. No one expects to see a new record here for many years, if ever. It is a fact that total manufacturing output has soared since then. The Industrial Production manufacturing index was 10.5 (2007=100) then and 99.5 in May 2014. That shows how huge the labor productivity increases have been in manufacturing. The US has the highest levels of labor productivity in manufacturing in the world and also the highest average annual rate of increase in this critically important measure over the past 70 years.

The construction sector was still down 1,496,000 jobs in May 2014 from January 2008. The government sector lost 507,000 jobs over that period, but almost all of these were at the state and local level.

Consistent with this shift in the type of nonfarm payroll jobs over the past 6-1/2 years, it should not surprise you to learn that the number of nonfarm payroll jobs held by women has been above the old peak set in February 2008 every month since September 2013. There were 68,393,000 nonfarm payroll jobs held by women in May 2014 or 49.4 percent of all such jobs.

As a corollary to the still-missing millions of construction and manufacturing jobs, the total number of nonfarm payroll jobs held by men is still below the old peak. It will take several more months to see a new record for men holding nonfarm payroll jobs.

Of course, there are two different measures of employment. In addition to nonfarm payroll employment, we have total civilian employment, which includes the self-employed and agricultural workers. This measure counts each person only once, whereas the payroll survey does not adjust for people who have more than one payroll job.

Total civilian employment peaked in November 2007 with 146,595,000 people employed on a seasonally adjusted basis. In May 2014 there were 145,814,000 people employed, so there are still 781,000 fewer people employed than at the peak. There were 9,799,000 people who were unemployed and looking for work in May for an unemployment rate of 6.3 percent. We should see a new record in the next two or three months. Then we can celebrate the fact that we are in uncharted territory by both measures.

The June 10, BLS report on “Job Openings and Labor Turnover” (the JOLTS report) told us that on April 30 on a seasonally adjusted basis there were 4,455,000 unfilled job openings in the US. That was the highest since September 2007, before the recession began.

The report also said that there were 55.1 million hires in the twelve months ending in April 2014. There were 52.8 million job separations in the same period.

Thus, we had 107.9 million people changing jobs over 12 months in order to get a net employment gain of 2.2 million people. The US economy remains the most incredible “jobs machine” every seen.


Dr. James F. Smith

Chief Economist


Share this: